TagRamp
AI-powered modern LinkedIn networks — connecting organizations from goals to meaningful business deals. Built solo since the World's Largest Hackathon (May 2025) by Bolt.new.
The problem
LinkedIn rewards posting, not pairing. Founders, partners, and operators want to find each other against shared goals — not scroll feeds.
TagRamp inverts the model. Organizations declare what they're hiring for, building, or buying. AI agents read those declarations and propose specific, actionable deals across the network.
What I built
- Goal-first matching. Companies state outcomes ("we need an FDE for a Series-B GTM rebuild") instead of broadcasting jobs.
- Agentic intros. A multi-agent layer reasons across goals, surfaces matches, drafts the warm intro, and tracks reply rates.
- Built solo on Bolt.new. Frontend, backend, agent orchestration, and ops — one person, one weekend cadence.
- Selected for Bolt 2025 — the World's Largest Hackathon. Public Devpost submission with demo video.
What I'm learning
Solo-founding an agentic product reshapes what "engineering" means. Every system decision is also a product decision is also a GTM decision.
Bolt + Claude + agentic patterns let one person ship a product an old-world team of five would call ambitious. The bottleneck moved from code to taste.